Skip to main content

Full text of "Energetics of tropical hummingbirds : effects of foraging mode, competition, and resource availability"

See other formats





This project would not have been possible without the 
help of tnany Individuals and institutions. I thank first 
Kristin Erugger for everything, including help with 
expesinental design, mist-netting, and editoclai review. 

To fetes Feinsinger I owe an incalculable debt, Pete's 
own doctoral research got me interested in tropical 
humningbicds while I was an undergraduate, and be has been an 
unwavering source of support, encouragement, and enthualasn 

on my committee have provided invaluable asalstance as well, 

the final product. My deepest appreciation gosa Co Martha 
Crump, John Eisenberg, Jaak Ewel, Haivey Llllywhite, and Jack 
Puts. Many other faculty and graduate atudsnta In my 
departmanc helped at varioua stages, including John Anderson, 

and Ken Nagy who helped me to learn the techniques for using 


I an graceful Co Che nany people vAo nade my rese 

nightnacee. In the field 1 waa ably assiaced by llaa ] 

Tim Keicc. Dwight Lawson, Cibdy Lordan, and Dcrug Kaaon 

Finally, I offer my aincereac chanka CO 
and Doris Campbell, Michael and FaCricia Fogden, Adrian and 

Pounds, Che Escon Aockwall family, Che Paul Smith family, Che 
Zuchowaki . 

My research was supported by the Depaccmenc of Zoology 

Florida, Mark SilUam aoppe, NSr grants to p. Felnsinger 
(BSR-8605043) and P. Feinsinger and H- Ilebout (BSR-81010U) , 
The Jease Noyes Foundation (administered through The 

cancer for Latin American Studies). 



assumed chat energeclc success was pcopoacional CO cates of 

First, I examined the consequences of sharing a feeder 

(sucrose solucion) , Birds tested in both conspecific and 
heterospecific pairs generally experienced 1 

behaviorally dor 

energy storage conrpared to solitary 
a heterospecific pairs than in conapeciEic 

, far) . when feeders were low>rewsrd, both 
species ware energetically unauccesaful at either distance, 

than Aoariiia. Using the DLW method, I tested the effects of 
feeder dispersion on heterospecific psire competing for 
limited food. When feeders were clumped (h 
aggressive behavior by Anaziiia restricted food int, 

territorialist . when feeders were dispersed (hence 

and pec visit intaSe ratea than the territorialist. 

R ^sneral; Chlorostilbon had higher eaergy expenditure 
than Anasiiia, due to higher rearing metabolic rates 
Thus, whenever intake wao limited, 
due to low-reward feedert or to interferenoe by Amarllia, the 

n appeared energetically speclalired for 
different oonditlona of nectar availability and ooogietltlon- 

Guilds df neotropical neecac-feedlns birds often contain 
many species and display rapid structural changes correlated 
with shifts in climate or food resources. Nevartneiess, the 

hunoningbird aasembiages <a} ahows chat populations trach 

proposition of my study: differential energetic 

consequences, due to speoiss-specifio differences in foraging 

foraging modes differ in their energetic costs and gains 

I focueed on several 
foraging costs and gains. Through 
I investigated (a) the flexibility 
physiological responses of individuals 

foraging sodea. 1 conclude ' 
syochesis of experimental results, discussing chair 

thought to affect 
of 5 experiments. 

Many field investigacioos describe conpositions of 
guilds (sansu Root 1S67) of both necterivorous birds and 

1912, iggO, Colwell 1973, reinsinger 1974, 1976, 1978, 1900, 

75, 1980, 1991, 1905, Wolf ec el, 1976, Des Granges 
ric-Brown at al. 1904, feinsinger et al. 1985, 1987, 

several distinct forsging modes that correlate with 
characceristlc morphologies. These modes, also celled roles 
or foraging strategies, may be differentially efficient ec 
utilising various food resource states (Falnslnger and 
Chaplin 1975, Wolf et el. 7976, reinsinger end Colwell 1978, 

FainBlnger ec al. 197! 
apacias in the Qulld i 
Trochilldaa {Stilea 1! 

populations occupyiog 
carrylh9 capacity and 

a typically reatricted 
1- Pence, diffetences 

, Montaomerie and Sass 1991). Final 

Species composition may chanpa over seasons or even d, 
such changes ace often correlated with fluctuations Ij 
food base (Colwell 1973, Felnsingec 1976, 1980, Gass , 
lertscian 1980, stiles 1980, Murcia 1997, Feinsingec ei 

respond to fluctuations 
DeohsnisDi(s) by whicl) m 


costs and benefits associated with food acquisition probably 
vary with fluctuations in the food base. If hummingbird 

efficiencies of food acquisition, then at a given reso 

ie«t tJiAir requirements more eesily than othere. 

n animal meets Its enerqy requirements {defined 

{ experiences "energetic 
• Is "eriergeclcally 

guild members may in tarn be 

flower distribution), 

that are energy stressed. Reproduction can be similarly 

changing structure of the guild may reflect the changing 
energetic success of each of its member species. 

density (Hiller and Gass 19d5f - Ragardless of the other 

Thus, although energetic success may not be sufficient to 
explain the presence of a given speciee, it is a necessary 

■d primarily in the laboratory, 

huriDinQblrda - Metabolic eApenditores of foraplnc humningbirda 
can depend on body alze (KacMillea and Carpenter Idll), ving 
diac loading (Eptlng and Caaey 1913, Greenewalt 1975), power 
for hovering flight <Epting 1975, 1980, aainaworth and Wolf 
1972c), altitude (Berger 1974b, 1978, Felnainger et ai. 

temperature (Seuehat et al. 1979, Schuchmann 1979, schuctunann 
and Schmidt-Harlob 1979a). Efficiency at extracting nectar 
may vary with bill and tongue length (Ewald and Milliana 
1982, Kingaoluer and Daniel 1963), length and ahape of flower 

poaaibly phyalologlcal conatralnta on food proceaaing r. 
Intraapecific coopetition among birda may alao limit fot 

and benefits vary 

rates of energy ei^enditure (i 
rhese studies indicate chat energetic coats 
loth with physiological and morphological 

vary markedly in chase chacacceclstics, species should differ 
in their energeclc responses to changes in Che food base. 

It is not surprising that so many environmental factors 
influence the enecgecics of hummingbirds. As a group, the 
Trochllidae have among the smallest body sizes and the 
highest mase-specific metabolic races of any endocherm, with 
the possible exception of the Soricidae, class Mammalia 
IJohnagard 1983] . Hummingbirds have one of the most 
energetically expensive foraging modes (hovering flight), 
relatively low thermal inertia, high surfsce-co'volume ratio. 

activity and maintenance coats, they most maintain high rates 
of energy intake relative to body mass. In addition, due to 
their extremely low energy storage capacity (see Paiadlno 
1989], to survive hummingbirds must be able to regulate their 
energy budgets precisely even over relatively short periods 

Recent investigations demonstrate the importance of 
energetics to foraging decisions made by hummingbirds. 
Studies of optimal foraging suggest that hurmingbirds follow 

forage witbln chosen patches ( 

e allowed to cbooae 
La appear to aelect thoaa flowers tfia 

Such choice experisienta alao suggest aevecal possible 

short-term energy intake, reapood to current energy 

energy eapended In foraging] (Gaea and Kontgomerle 1961] , 
Enerov Wanageipentr Tine nuOoer.e and TerTlrnrv SlTe 

Field studies on territorial species indicate that 
hummingbirds guickly adjust energy intake relative to 

1963, Hiron et al. 1963, Paton and Carpenter 1964). Time 

RainswQcth and Wolf 1972b), and Indicaea that huiruninghlrda 
are able to adjust behavior rapidly when competitor pressure 
and food availability change. Extrapolation from time 
budgets suggests that their energy budgets nay be Influenced 

Throughout my study I 
to denote the behavioral an 
individual that deterroine 1 
expenditure, and storage; a 

physiological responses of an 
9 total energy budget (intake, 

cerm avoids several problems inherent in oommon expressions 
such as "maintaining energy balance" and "balancing" energy 
budgets. First, these expressions Imply that there exists a 

considerably. The idea of balancing a budget is also 
misleading, because energy budgeta are expressed as eguaticns 
and ate thus by definition balanced. Eecause energy budgets 

energy management or 

basal metabolic race) no 
ore neutral expressions 

Much evidence suggests Chat hummingbirds should be 

variety of potentially energetically stressful conditions. 

Keveccholeas, to date no study has directly oiaeaured energy 
budgets o£ huiMiingbltds foraging under controlled conditions 
of intruder pressure and food resource states. This study 
foeusee on five related determinants of energetic success: 
foraging mode, food processing constraints, competition for 
food, food availability, and food dispersion or density. The 
experiments discussed below evaluated tne energetic 

Because many other factors can contribute to 
energetic success, I sttempted either to hold these factors 
constant or to control for them by using rastched-pairs and 

In successional habitats at the lower elevations In 

the course of a year (Feinsinger ISIS, 1576, 1977, 1978, 
1980, H. Tlebout, unpubl. data), representing a wide variety 
of foraging modes (Feinsinger and Colwell 1978], Two modes 
that differ greatly io Isoth behavior and morphology are 
territoriality and low-reward traplining. The following 

primarily from Feinalnger (1978), Feinsinger and Chaplin 
11975), Feinsinger and Colwell 11978), and Feinsinger et al. 

other hunoBingblrtis, 

d hummingbirds. 

In terrlcotlalis: 

ind meneuversbllity. Low-seuard 

! excluded by resident territorial 

tar-feeding guild in 
s iFelnainger 1916). 
ranted HuiBtoingbird, i 

i trapliners have very 
relatively low 
d forward flight. 

species (e.?., Inga bceitasiS, Leguminosae; Hasielia 
a has Blph win? disc loading (.0335 g-cis‘*, Felnslngar 

species la Cousd ac the study site throughout the year, with 
peats In abundance during March through Hay and septeoiber 
through Novenibec (Felnslngar 1976) , ChlorosCllhon caniwecil, 

are widely dispersed when 

reward flowers not often visited by other birds (e.g., nuhus 
spp.f Rosaceae; Acnistus arborescens, Folanaceae; and Cuphea 
sp., Lythraceae) . Chiorostiibon has a short (ca. 18 mm), 
straight bill and low wing disc loading (.0262 g.on'f, 
Felnsinger ec al- 1979). A ZS% difference In wing disc 

elevation of Honteverde (funazilia > ChlorostiZbont (Felnsinger 
ec al. 1979) . Chioroscilhon is found In Honteverde frois 
January through November, with a pea)c in abundance from April 
through June (Felnsinger 1976) . 


together ec the study site for noat of each year, the two 
apeciea iateraot fregueotly In the field — especially at rich 
nectar sooioea. Pelnainger 11976) obaetued that Anasilla won 
991 of aggressive encounters with Chlarostilban. Considering 
aggressive encounters anong all guild members, those by 
Cblo^ogtilben ware primarily interapeclflo, whereas those by 
Aisariiia were mostly intraspecific. Felnslnger concluded 
chat, in terms of population- level CMspetltlve effects (sensu 
a relatively large impact on 
I, primarily by excluding It 
from rich nectar sources, whereas Chloroatilbon had 
relatively little effect on Amasliie. In addition, Amsriiis 
could Unit their own local populations through Intraspeclfic 

oonpsrs-ive study of the tour addltionel deternilnents of 
foraging profitability presented below. The above 

, coupled with their divergent foraging modea, 

o species should differ markedly In their 
c responses to these 

CAloroaclibsn when food sources are rich and defensible. In 
contrast, Chloroscilbon^ with its low wing disc loading and 
exploitative mode of competition, should face better than 
Amariiia when food sources are limited or widely dlepecsed. 

In addition, because the generic pair amaailia-Chlorosciibon, 

geographic range. 

PhveiQloglnal Con^rralnri on Ratet of Food Proce^alng 

Chat huniDlnghirds la the field may typically be limited in 
their food Intake ratea because of physiological constraints 
on food prooesalng, pcicnarily digestion and absorption. IE 

phyelological 1 
J quantified fo. 

relative importance of 
the two study apaoies. 
h species following ad 

then compared my reauits Co theoretical v. 

With their high energy EeguiretnentS/ eenaitlvity to 
energy screes, and diversity of foraging nodes, huirutingbirds 
are an asceilent group in which to investigate the possible 

categorized inC 

s of conpetition, interference ai 
a recognize as cnany as six kinds 

nodes utlllalng the sane flower species (Stiles and wolf 
1970, Wolf 1970, 1970, reinainger 1974,1976,1973,1990, Stiles 

Gill 1978) . To date, however, the actual energetic 

Each of the two foraging modes I studied corresponds 

1. Tieboot pers. 

patterra (Fainaingar 1976, H 
terrltceial behavior, such ae calls anb dlaplays from a 

Many others, however, appear to be as costly 
core so, as when the territory resident chases 

In contrast, low-reward traplining hummlngbicda, snch as 
CftJoroscllbon, typically visit widely dispersed flowers and 

enter Amazille terricorles, but Chlonatilbon do not initiate 
aggressive encounters and usually leave flower clumps 
defended by resident hmssille. Thus, species with a 

ejtploitativB competitor (Pelnslcger 19761, or consumptive 

Interference also compete by exploitation, in the sense that 
available food must be divided among participants. In 
addition to laoilng interference, exploitative competition 
may also Involve benavlor le.g,, food location and timing of 
visits) that increases food scguisition relative to others 
using the resource. For example, trapllnera may increase 
visit rate to flowscs and harvest nectsr "pre-emptively," 

According to Folnslnger <1976), 
uch a aklllful crapllner of dlapersad 

ahort-tecm bahavlocal reaponses and anargatic C0A9a<Tuancae of 
n eeaily defanaibla food soucca. I tascad Anaaliia 

llbicupi faedac [an artificial flower containing unliaiitad 
sucroae aolution) Lindac thraa conditlona: aolitary birda, 

ganaral goaationa. (1) Do hummingbird apeciea with different 
foraging modea axhiblc different behavioral and energetic 

pair, cooapeclfic or heteroapeclflc? 

Daaplta the preaence of behavioral Interactiona between 
paired blrda that indicate competition, one cannot asaume 
that theae interactions will have energetic consequences. 

birds share food (e.g,, terrltorlai display flights) might 
courtship display). Thus, although there nay be time budget 


food Mse can be of prime enec^sclc importance to foraging 
humningblcda: food availability and foraging coat. I define 

eitbee ao net anecgy gain (energy intake - foraging coat) or 
aa foraging efficiency (energy intake/foraging coac) . One 

oea foraging cost, and hence 

Similarly, at a fixed incerflover 
distance, as flover reward decreases, foraging cost Increases 
because a bird needs to visit more flowers to get che seme 

does profitability, birds foraging u: 

maintain optimal foraging profitability. Conditions chat can 
result in decreased profitability subject birds to "foraging 

n energetic 6i 

availability In otclec 
previovaly defined. 

Ron do buirolngblrda respond to noncorg>Btitlve conditions 
of foraging constraint, In vbich flowers either have low 
nectar reward or are widely dispersed, or both? Because 
species with different foraging modes nsy differ in foraging 
costs and forage most profltaBly at different levels of food 
avalleblilty (relnslnger and Chaplin 1975, Wolf et al, 1976, 
Feinslnger and Colwell 1978, Pelnslnger et al. 1979, Brown 
and Bowers 19851, they isay respond differently In both 
behavior and energy management- The goal of ray third 
experiment was to examine the short-term responses of 
hummingbirds to experiments! conditions that simulated 
nature! foraging constraint. 1 addressed two specif lo 
questions. (1) What are the behavioral and physIologlCBl 

foraging constraint? ( 2 ) hre there species-specific 

a energy management that relate to foraging 
d solitary individuals of each species in a 

Competition for food can interact with food dispersion 
terns to affect foraging profitability and overall energy 
gets. When nectar is limited, competition for food can 

affect energ/ btCgeta In several ways, first, intake stay be 
affected because available food is divided among gompetltora. 
Second, foraging costs per unit nectar consumed may Increase 

depends on the types of competitors pr 
dsfensibllity of tne food source. 

energetic coneequences of eharing llnl 

method (validated In 

t behaviorel and 
1 food at different 

(eesily defensible) and widely dispersed loot easily 
defensible) . My goal was to cast the prediction thet 
Ajoeziiia would be relatively more euccessfuX energetically ai 

better when feeders were dispersed. 

n Chapters 3 through d 

and predictions to be test 

experlmentel conditions 

only. In chapter 8 I Intograte tho results of the 
experiments and suggest extrepolatlons of my findings 


s at Konteverde, c 

aupplentented with 

optical* model 10 

I chanoed food anc 
birds eapexienced 

ormlijaal aviary on fluorosa solution 
avian vitamins and Orosophiia, ad JibiCum. 
solution was made with locally obtained 

solution was 201 sucroa# by weight (± 11 ) 
and-held optical cef tactometer {American 
ill, preolaion ±0.251 (-0.25°Bri!(] ) , I 
1 fresh water foe drinblng and bathing, and 
water daily. Throughout their captivity, 

1 pbotoperiod {approx. 12 L: 12 D) and 

outdoors as possible (indoor mean daytime temperature 21-27 
•C, relative humidity up to 1001). I mar):ed birds 
Individually with enamel paint and colored acetate lag tlaga 
1 identification. The 

1 -Waive raity 

Animals, University of Florida, approved a. 

Birds stayed in the comnunal aviary E> 
to being used in experiments, to a 
captivity, 1 moved birds Co individual holding cages 1-1 > 

received no ProsophJia, Examination of excreta (urine and 

birds to clear all Prosophlia from the gostro-lntesclnal (GI) 
tract (Halnsuorth ISIS). During eaperimenta, 1 fed birds 
20.0 to. 251 sucrose solntion made with fully refined table 
sugar purchased locally and gave birds no Drogophile, This 
diet produced clear, colorless excreta that were apparently 

bird In only one experiment, except that several birds In the 
DLW validation had also participated in Che previous 
experiment. Following each experiment, with en eye-dropper 

The standard equation for e daily energy budget (after 


efficiency. Ke aieo found that othi 
couXd represent energy lose, euch ai 
praacnt in appreciable quantities (- 

To judge a bird’s energetic ai 

ccess requires knowing its 
preparing for tnigrationr for exsisple, Che preferred 24-h 

pcejDl?r3cory or pcereproduative 
budgets that result in mass gain 

preferred energy budgets eas to test birds under conditions 

did this experimentally by offering birds ad libitum food at 

compared the preferred energy budgets measured under such 
"control" conditions to budgets tested under various 

I {e.g., food limitation, competitlonr 
le In time as possible to tbe oenttols. 
control energy budgets, Z assessed 

treetment group (s) with the highest P were the least 
energetically stressed, I justify using P as the best single 

First, P combines Intake telatlv# to expenditure in a single 
value. Second, it is the beat energetic predtetot of how 
long an animal can aurvlve should anergy Intake be 
interrupted Isee below) . Third, there are no circumstances 

under which birds would maintain P rates higher than thalr 
preferred level; thus, when comparing Pa among different 
treatment groups the highest can be assumed to be the closest 

scale as a fcnctioa of body nass 

hence laecabollc scope} ulclcssceiy depends upon che ability o 
a given unit of oiecabollcally active tissue to petfocm work, 

specific units reflects the organism's ability to meet both 
tissue-level output raqulramanta (expenditure) and supply 

demands <lnta)ie) (Hainswortb 1981a, p. 166 ) . 

expenditure <e.g.. bmr a mass*^*, see Nagy 1987). Just 
because scnsner birds are predicted to have higher mass- 
specific tnetabclic rates Is not a juatlficetion for removing 
this difference, as would be accomplished by expressing 

relationship (a.g., J -h'l) . The differences in mass- 

specific Dietabolic rates that are due to body size reflect 
important differences in surface area-co-volume ratios (Brown 

similarly, axpraesinq ariargy terma on a p«r diera whole 

1 calculated energy budgete uaing the OLW nethod 
(described in Chapter S) and the standard gravlnietrlc method, 

I • [<lrMfSUC]M16.4B)l/!(SM) (T)I (2.S) 

; of energy scorage from dif 
>dy nvasses recorded before eod after 

1 a cop-loading electronic balance (: 
Carpenter et al- (1983) , Thi 

before recording body maea. The balance pao was covered by a 
small plastic screen to ensure accurate readings, unaffected 
by wind and excreta. The digital display on the balance had 
a stability indicator light that denoced when a reading was 

cause uncertainty In laasa decerminations . Each body mass 
meaaueetrent followed a food deprivation period Chat was 
sufficiently long to allow passage of excess weter (see 
Chapter 3 for validation of this procedure) . 1 assimed that 

all changes in body mass represented losses or gains of fat, 
for both daytime mass changes (following Hainsworth ec al. 
1977) and for 24-h mass changes (following Hainsworth et al. 

apeciea of roosting hnawingbirds (Kruger et al. 1982) support 

this assumption. Thase wor)(ers found that, starting about 10 
icin after birds stopped feeding, S^s began to drop from 1.0 
(pure carbohydrate metabolism) to 0.8 (pure fat necaholism) . 

((BMr - BHi 

).9) l/[ (BM) (Til, 

-here P Iptofluetionl Is scotafl ensrgy |J-q-i -R-i) , aHe ana bMi 
srs final and initial body mass <g), Esspectively, and the 
heat of oonibuatlon of fat Is (9-5 )tcal/9)(9.J tj/kcall - 39.9 
W/y (from Sehmidt-Nlelaeo and Sctmldt-Nlelsen 1952, Scnmldt- 

othet than fat le.?., ylycogen, protein, or blood glucose), 
then my values of p would overestimate energy storage. I had 

systematically bias one treatme] 

I calculated energy expenditure f 

I also talcolaced a Proficahllity Index (PI) as 

lar to "foraging efficienoy" (e.g,. Wolf et al 
because it includes non-foraging expenditures 
r indicator of overall energetic success, I u 

>1 efficiency, Bligh ai 

During each cbsecvatios period X recorded aev 

particular aaperlnent and lea deaign. I preaenc m 

acore Chen. Exceptlona to these criteria ere notei 
Metbcda eeccion for each experiment. 

visits to feeders were tallied for h 
period and auinried for all periods ccrabint 

during a foraging flight (perch-to-feedei 
while hovering in front of a feeder, a bJ 

addltionei vleits to the seine feed. 

presented as rates (vltite/hl . 

d Dy dividing tl 

he feeder aperture 
ider (during the same flight) 

corded directly from an 
feeder) following each 
e tocal inta)ce par period 

a-held stopwateh (±0.5 s per flight) a 

included both hovering end forwerd flight, 
although thasa may entail dlffacent enacgetic costa iG: 
1S65) . The tine apent flying is praaenced ei 

When birds ware in pairs, I recorded agonistic bahavior 
believed to be energetically important {Ewald end Orlana 
1983), t.e., interactions in which at least one partlolpant 
was in flight. I defined three types of interactions as 
follows, with the focal bird sooted as either an Initiator 
(■h, initiated Intetaotion by moving towards eagemate) or a 

At . Initiator and receiver roles never swltobed during an 
interaction, thus, the initiator always '’won" the 
interaction. (U Chases occurred when A flew towards B while 
a flew away Irojii A. If) Body contact occurred when A flew 
towards B and touched any part of B- When birds flew towards 

infrequently) body contact was not scored. {3) Perc) 
displacements occurred when B flew from its perch inni 
following body contect or a chase initiated by A, or 
following a vocallretion by A. A then flew to occupy 
perch locetion previously occupied by B. Similar to t 

single flight were i 

multiple interactions during a 
hen the initiator fed 

contact oocucteC within a pcaCefinact distance (apeciflad for 
hehavioc that occurred further chan the predetecitined 
chases. 1 used feeder defenses as a distinct category of 

encounters further fron feeders. All agonistic behavior 

birds (bill fencing or body jabs) that did not involve at 

determine which bird of a given pair controlled the feeder 
for each period. The bird initiating the most acts vao 
scored as winning control of the feeder for that period, with 

cagenates initiated the same frequency of acts. 

Analvaea and Statlatii 

Before analysis. I tested all dependent variables for 
homogeneity of variance using Bartlett's Test tSystac* STATS 
module. Wilkinson 1981) and then used appropriate 

z effects ualn^ 2 oc 3-way ANDVAe (Systat® 
HGLH nodule, NilAinson 1997), depending on the experiment. 

variation foe moat measures, tbis design greatly enhanced the 
ability to detect group and treatment effects. 

I refrained from using multiple comparison 
t all possible pairs of reoacLS, as they are 

Instead, I tested only a small aubset 

standard and repeated-measures AKOVAs (Day and Ouinn 19891. 
Significance levels reported for factors or their 

Interactions are from cfie ANOVA, wl 

Linear relationshlpa between variables were tsateO using 
the Systat^ MGLK module <Hllkirtson Isa*?} and BMDP So rusdule 

correlation coefficient (reported as "r” in other sources). 

I report R for both correlations and regressions, because the 

e reported f( 

determination, zar 1984; p. ZIU. ! 
tests of significance, and I report 
8.10 as trends (Siegel 1956, p. 9} w 

ri, the coefficient of 

5, and 7) to provide results applicable to natural systems. 


by many laboratory ecolo^iata: 
' experimental reaults be 
.e wild? Because of the large 
1 conditions that 

for the system I studied. 

To address this concecn, 1 structured my experiments to 
enhance the applicability of my results. Bach experiment was 

magnitude of responses le.g., dmssills < Chlorostilhon by 21t 
for energy expenditure) as well as their absolute magnitudes. 

relative differences have a greater likelihood of persietLog. 

distance x food level), while na)clng interpretation of 
results more complex, allowed testing for statistical 

simultaneously to determine the energy budgets of birds in 

clearly illuscrates Che 
factors, I employ a graphical formac. 



d intake rate in Aaaeiiia and Ctiloroatilbon is cncmally 
:d physiologically by their food procesaing cates. 
Diamond et al. USES) and Kacesov et al. U986) have proposed 

Aucnminphirda Caiypt^ enna and Selaaphoevs ru^as. Second, I 
determined if passage races in the acndy species are reliably 

for birds following reasonably short periods of food 

By examining the rates at which food empties from Che 
crop and is assimilated in the intestinal tract. Diamond et 

general model of digestive c- 
hunmingoirds. First. 

negative exponential fnnction of time. They then proposed 

acidification of Che meal In the stomach, and absorption of 
sugars in tba small intestine. Finally, they suggested chat 

these physlolosiceL coaetselnte on food peasage rates limited 
rates of food intalce- Thus, the large amount of time 
huEmningbirds spend perohlng Instead of feeding iStlles 19'ii. 
Wolf and Ralnsworth 1971, Ewald and Carpenter 1978, Risen et 
al. 1989) may be required for the crop to empty suffloiently 

limited In their food intake by physiological processes, I 

Che got sec crop-emptying time, as Karasov ec al. proposed, 
then passage rates through the GI tract should reflect 

In appreciable amounts, U) excretion races should folloe a 
negative exponential decline over time. If crop emptying 
follows an exponential decline over time following a single 

passage from the crop may be a positive function of Che 

crop emptying time end lb) digestive constraints 1 

maximal crop emptying sates land thus maximal excr 
races) should he poeltlvely c; 

passage rates usually limit food lnts)ce . 
al, suggest, chan 13) maximal e 

rates they wsold sot be pbyaiolaglcslly c 

species CD (a) deeding rates of nonnally aotire birds in the 
laboratory and (bl perching times in both laboratory and 

n yield reliable stable body masses without undue st 

I measured rates ot food intake, excretion, and the tine 

0.19 ql . Cara and feeding of Oirda are deacrlbed in Chapter 

individual cages 1 n^ with ad liOitum 20. Ot sucrose eolution 
(but no Droaophlla) for 2 h, X then deprived each bird ot 
food for 75 min while recording its body mass from a perch- 
balance (± O.flOl g) every 5 min (the “Ad 1 runs) . This 
length of time enanred that the crop and Qi tract would be 

To ensure an adequate range of meal sites across the IS 
trials, I varied the contents of the vlsl from 100 to 500 pi, 
1 weighed each vial before and after feeding, and recorded 

for another 75 min while recording body masses every 5 min 

i« end of Che focaglns 
sC prior to rooecing. The protocol was 
Eoeal {offered at ca. 1800 h, the normal roosting time), 

changes in body mass prior to reaching a stable level 

guotlent IRQ), or all of these. Because I recorded body 

was probably oonstant at approximately 1. Birds consumed 
more than sufficient water to compensate for EWL and maintain 

h excretion, , 

iurlne and feces combined) , 

> evaporation, it would be e 
: analysis, 1 included EWL x 

I analysed data from each run individually by nonlinear 

wtiexe Htfc ■ body mass ioas during tbs 5 min period ending ac 
HC, He ■ tine (min) since deprivation began, and PI-P3 » 

unbnovn paraneters solved by Iteration. 

50 to 75 nin) . Then I calculated the time at which the decay 
curve <eg. 3.1 for each run) intersected the 955 confidence 

reach stable body mass. 1 used the teaxiteum mass loss during 

respectively) t 
excretion rate. 

1 performed statistical t 
ANOVA (BMDP program 10; in cas< 
the Brown-Foraythe AMOVA — Oixoj 

1 nighttime trials separately. 

exponential functlona. 
response as soon as deprivation 

.iise) before a peak followed by exponential 

I were highly positively oorrelsted with meal 
(Table i-I}. In all oases the coeffloient of 
Increased with Increasing tine period. 

> ad lili runs, maxS did not differ significantly 

e distribution 

Anaeilia and Chloroaellbon, respectively) . 

right, with no KAX5 values greater than 5S mg (fig. 3-2J - 
Within each species there was no significant difference in 
the maximum MAX5 value for ad lib (IIB vs- 56 mgfS min, 

meal ruhs (125 VS. 55 mg/5 min, Ameziiia ai 
respectively) . In each case Amarilia exhi) 

between meal sire in mg (HStZE) and < 
deprivation to reach a stable body m. 

e relationship 

TSTMLE - 0.09711MSI2E) ♦ 

Ollorostllbon-. TSTABLB • 0 . lOOl (K5IZEI * 

it was iispaaaibls to calcolats tstasle for socstlog bi: 
"stabllizsci," itioBt of tne watai fron tHs last meal was 

± 10.9, /maziliit sob CbiorofCXlbon, respectively) end : 
□o significant correlation with the else of meal eaten 

proceaeea aet crop ampcyiAg races, whlcb also conform Co a 

pcsdlccion 2, Is also conaiscenc wlcb a negative exponential 
nodal of crop emptying but not vitb a linear model- This 

part of prediction 2, which was derived from the suggestion 
that digeacive processes per se set crop enptyiog time 

tract, because if processes in the GT tract constrained crop 
enpcyiog races, then there should exist a crop volume 

as 297 pL (ChioroscUbonl and 372 pi (Amasllia) . This 

ABsamlag that caaiciinup s. 

flguies agree closely with experinentaily measured ad iihitcrsi 
ChJorosciifioh, respectively (see Chapter 4) . Using the 

If these latter estiinatcs represent the physiological 

maximuji] potential cate. I. 
CtlloroSiUbon generally si 
Araeailia (Chapters 4,S, an> 

holds tree in the wild. To exploit its typical array of 
widely dispersed resources (Feinsinger 19T6, and Feinsinger 

waking time in foraging flight (P. Felnalnger, pars, cosim.r. 
This is consistent with flight tines reported for 

birds nay actually n 

closer Co CCelr physiological limits o 

positive relationship between flight t 
the relationship betveen sitting time i 
constraints on feeding rates seems teni 

itinimlaihg energy expenditure 
hummingbirds . 

do not incur an irreversible energy debt (see Tot 
19G5) . Hainsuorth and Wolf U972a) found that fj 
through a humnlngbird's digestive tract in about 
Accordingly, HainsworCh et al. (1977, 1991) used 

1985), and one ctloroatilbon had to be revived by warming and 

« regression equations for MSIZE vs. T5TABLE, I 

still react! a stable body mass wlchlo 30 Bin. fotentiaUy, 

20) sucrose solution (eg. 3.Z|, compared Co Its mean meal 
alre of 112.1 1 42.0 mg (the solitary control days from 
Chapter 4) , Chlocoscllbon could process about 13S mg (eg. 

solitary control days from Chapter 4) . If crop contents 

they eaceed twice the average meal slse, these figures 
provide Independent theoretical support that stable body mass 
should be reached within 30 min after deprivation. 

Finally, because passage rate is a negative exponential 

errors in estimating true stable body mass. For example, one 
AfliBxlIia during daytime deprivation after ad libitum feeding 
reached stable body mass in 15 min, after losing 74 mg. The 

closely with the figure reported by Karasev et ai. (1366) and 
Diamond et al. (1986), based on a negative exponential model. 

solution from the crop (a 100 pi meal consists of 86.4 mg 
Roosting Axtasills and ChloroeCiihon processed crop 


celicble stable body 

(Haloaworth et al. 1991) . ' 

d LVttponis clemencias 

eneE^etically cnoce efficient. Kzuget ec al. {1992)« howevs 
daytime valuee of 1.0 {cacbohydtace metabolism) to 0.8 (fat 

aesve to protect aqaibat excessive evaporative water loss 
during tbe night . 

From this experiment X conclu 
Chloroatilbon are digesta-free. Thus, changes 1 

noimal rates of feeding, even wheo Dirda take large neals 

are pzo&ably not limited by digestive processes. 
CblorofftLlbon, hovever. may be closer to reacbing lbs 
physiological limits than is Amazilia. 

Correlation between meal site 





Figure 3-1. Repreaentatiwa t 

■ Ad Lib njns 
G Single Meal runs 

1 Amaillla saucaroltal 


lilini ni 

O-S le 31 32 10 IB SB 61 73 BO SB 96 1Olll21!0’2S 


Figure 3-2. Maxioiuo body ness loss in 5 rain (MAXS) follawiog 
ad libitum feeding and Eoiloving single meals during daytime 
approximately normal/ while chat Cor chloraatilbun is sharply 



Few stediee have succeasfvlly raeaaured the enezgetlc 
costs and benefits of competition In aninvais. The oetabolic 
expendituces of interference competition have been directly 
determined in several ectotherms (Bennett end Houcle 19E3, 

Jaeger et al. 1965, Rieohert 
Arnold (19E2) reported high r. 

e competition. No ( 
f exploitative competition 
n energetic costa or hi 

oompetition for endetherms, althoogh Roaliaft et el. (1386) 
and Hogsted (1987) report high metabolic rates for indivlduel 

flocRs, and Malloy and Herreid (1979) report elevated 
metabolic rates for fighting Mjs mcscuius, 

Buouaingbirds provide an ideal group in which to 
investigate the energetics of costpeticibn for food becsuse of 

on A honognneous diet of known energy content: sucrose 

measured, then ite energy content can be precisely quantified 

humraingbicds consume food in discrete nesls, without 
aeleocivlty (removing preferred items fcom within a single 

easily be determined. When artificial feeders are oriented 
vertically with e feeding aperture on top, birds rarely spill 
sucrose solution. In my studies, close examination of birds 
using such feeders established that little if any sucrose 
solution adhered to bills, as was confirmed by the near 

characteriatica. when birds shared a aingle feeder, I could 
determine energy intake for each bird accurately using 
gravimetric techniques, Pocd competition experiments using 
gravimetric techniques are difficult to perform with animala 
ause spillage, aelectivity, and 

foraged at dispersed flowers when it was excluded f 

ocBl populations of 

interactions used by Asiesiiia and CMorosCiibon 

possibility, my second expeciiiient was desi9ned to determine 
the behavioral and energetic responses to sharing an easily 

conditions of competition not normally f 
example, an intruding cblorostilben mlgh 

necessary to allow detection of subt] 

consequences experienced 
pecicds in wblcb cbey actually share a rich food source. 

specific responses for Anaeilis and Chloco3ti2bon chat could 
relate co foraging node, and {b} determine the relative 

Che folloving predictions for conspeclfic pairs of AmesiJia 
(hAI ano ChlorosCllbon (CC) and heterospecific pairs (AC) 

aolltary individuals. Recall that Ajvaallla behaves primarily 
as an interference conpetitor, whereas Chloeostllbon behaves 
pritiarily as an erploltativa conpetitor (Chapter 11 . I 
predicted that both species w 
hovering when in pairs than w 
flight cine could result tcon: <al the crapllner 

frequent feeder visits (both CC and AC) (Chapter 1, also Sill 
I9aS) and flying more due to being chased (AC), and (b) the 
territorial Amarilia competing by interference, hence 
increasing flight time due to chasing and being chased (AC 
and AA] , Because novering and flying are the moat expensive 
activities far a hummingbird, I predicted chat birds in pairs 

feeder would be asymmetrical (Perrson 1905) . In CC )iairs, I 



eCCecc on anergy storage, in aa pairs. I expected that 

cates of energy stozege zelative to contcols, Finally, in AC 
pairs. I expected that ChiorostilOon would act he able to 

resulting In considerable reduction in energy storage for 
Chiorostiibon in heterospecific pe 



factors: spscies 

followed by ac aaporlcLOctal day (paired birds) . T conducted 
six trials with heterospeoific pairs and three trials for 


in coupeclfio pairs and Che ocher half cesced first 
hecerospecific pairs. Psicners were randomly assigi 
bird apenc 4-fi d hecween crisis on sd libicorr 20S st 

loadihg eleccronic balance {±0.001 9). AC Che scart 

and Che birds shared a einqie ad iibittojn feeder loct 
between cheir perches and attached Co a cop-loading 

defensible from either perch. The protocol for eact 

feeder(s), and birds 
sin four times during 

deprivation period 

boring each of the five feeding periods (0530-0730, 
respectively) I recorded visits co feeders (1-h observation 

enacgetlcally; Caeder defenses end agonistic e. 

(defined in Chapcec 2) . feeder defenses were scored only 
within 25 on of the feeder and were recorded during each of 
the 1-h observation periods in which I recorded feeder 
visits. During the SD-niin observation periods in which I 

further than 25 osi frois feeder: perch displacements, chases, 
and body contacts. I also cslculatad tocsl feeder defenses 

each ewperisiental day. These totals, which were the seme for 
the total nuirber of interactions 

(deEined in Chapter 2). 

On control days I 

near cagea daring each trial) 

PI) gravimetrioally (aee Chapter 2) . For all terns 1 us 
he sane body mass for a given bird: its initial mass at 

s deacribad below. 

Itacfta of ovfirnlgtit expenditure were caieuleted from t 

raoetinq, and BMp ia the final body maaa (9) , recorded at OdOO 
the followlnq oiorninq. 

1530] or 24-n (OSOO'OSOD) bloclee. The daytime block atarted 

fclXowlaq perlode one and four, raapectlvely. Period one waa 
not Included because aubjecta nay have had eapeclally low 
body eater levela when the body erase was recorded at 0500. 
Period five waa excluded because rcoatinq birds do not reach 
a reliable dlqesta-free naas within 30 irin (Chapter 3) . The 
period 2-4 nasa ceasurementa were taken under identical 

energy budgets. Por sotne daytime results involving only 
behavioral meaauces, and hence not subject to the sane 

species as a grouping factor, pair type as a trials factor. 

afifi bird nujnbec {alidce • control day vs. pair ■ experimectal 
dayi as cepoated measures on a slnple subject- other tests 

Sollrarv RH rdn 

On Che average, Antasfiie end Chlorofftilbon exhibited 
very similar behavior under control conditions {Table 4-1) , 
Although Amasiiia took significantly larger meals, both 
species had mean meal alaes that uera approximately 3» of 
their respective body masses. Both species spent about one- 
fourth to one-third of their daylight hours in Che air, 
feeding about every 7 min- for each species, incerbout time 

time {Table 4-2) . There vas no significant relationship 

Control birds exhibited slgnlficenc interspeoiflc 
energetic ditferenoes- During the daytime, ChiotostlJhon 
expended energy and coosumed food at greater rates than did 
Amsriila (Table 4-3) , Although expenditure rate varied over 
tuo-fold among individuals, birds were able to maintain 
daytime lnte)te rates at or above rates of expenditure (rig. 
4-2). As a consequence, energy storage rates (Table 4-3) and 
profitability Indicea (Table 4-4) indicate that both species 
maintained equivalent net profits during the day and exactly 

{Table 4-5) . The regiessiona of daytime expendituce againec 
hover time foe each speciee {Table 4-Sl were ueed to teat 

Birt-Frleeen (1985) . while Doth speciee h. 

conspeclflo pairs (18-2 vs. 50.2%, 9 < 0-05. AiMllJla and 
Chloroattlton, respactively) . Among palrad birds, Che time 
spent hovering varied an order of magnicude. The individual 

Lr (8.0%}, while Che highest mean h 
jstilbon in a heterospeciCic pair ( 
differad by spades within 
[bon taking signifioantly smaller meals than 

(? ■ 0.088}. This indicates that, relative t 

controls, each apecies iocreaaed mea 
hecerospccific pairs, but decreased 
Gohspeciflc pal 

meal siaes during hecerospecific trials, with an Amasiiia 
caking the largest (310.5 mg) and a ChJorostlJbon the 
smalleac (31.1 mg] . The single lacgeat meal conaumed by an 
individual in each experimental treatment group (Table 4-8) 

was always ac, or in excess of, Che cheoi 

IhCerbouc time showed a slgniflcanc three-way 
Interaction among all factors (speeiea x bird numbe 

d frequency In conspecific pairs. CAlorostilbon 
e opposite response, resulting in a considerable 

in riean intecbout times £ 
from i . S ±1.6 min (conspecific pains) to 16.7 ±16.4 min 
(tietenDSpecific pairs) . Interboue time was significantly 
positively eorceiated with meal slse for CiilerostilSon but 

Ability to control tbe feeder differed by species and 

conapeclfic pairs. Anazilia tied for control of the feeder as 
frequently as they won or lost. In contrast, Chiorosclibon 
almost never tied in conspecific pairs. In heterospecific 
peirs, Araarllla non every period over Chlorostllbon, with the 
esception o£ one tie in which each bird Initiated one 
agonistic act. 

Frequency of initiation of both feeder defenses and 
agoniatlo acts exhibited the same pattern (Fig. 4-7). For 
both measures, Ameallls was significantly more aggressive 

2 ±59.5 VB. 5-6 ±10.4 agonistic 
species showed relatively low 

n, loean o£ both species) . When 
>evei, Anaeilla increased its 

.ero. Aaazilia was capable of 

than Chloroatllton (e.g., 3< 
acts/h, respectively) . Bott 

in haterospecific pairs, he 
aggression levels to nearly 

31 agonistic acts/h. 

of aggression. The 
individual Amsailia 
r conspecific and 

heterospeolfic pairs, respectively. Despite its relatively 
low mean levels oJ agooletlo acts, CliJorostlJBon was also 
capable o( occasional periods ot Intense aggression, 
particularly in conapeciflc pairs. Single bird mctrsmes were 
96 and 46 agonistic acts/h for conspeciflo and betaroapeclfic 
pairs, respectively. 

1 examined Che relationship between hover cime and each 

time decreased, such chat at the highest aggression levels 

Entrgeiics. As a gronp, birds ot both species 
significantly Increased energy expenditure when in pairs, 
without dlstinctlnn between pair types IPig. 4-9) 
CMorosciihon maintained daytime rates cf energy expenditure 


(247.7 because Arasaliia increased expenditure In 

pairs only S.3», whereas CftJorostlJbsn increased 19. 8» in 

Species diCfered in cheir ability 
d expenditure by increasing foe 
Chiorcstiibon naintained higher inca)ce 

In addition, there was a highly slgnttioant interaction among 

> < O.Ol). THis interaction indicatoa that Chiorostllbon 
increased intake in conspecific pairs, but decreased Intake 
in heterospeclfic pairs, while ajxaztija responded in the 
opposite manner. Compared to control birds (Pig. 4-2), 
paired birds had difficulty maintaining daytime intake above 
expenditure (fig. 4-11). One fourth of all paired birds fell 
below Che line of intake-expenditure equality, including 

The net effect of responses in expenditure and lnta)te 
was that birds in pairs had signlfiosntly reduced rates of 
daytime energy storage relative to controls (p < 0.01; Fig. 
4-12). Amaslila in heterospeclfic palra suffered the least 

reduction in daytime see; 
±65.6 J-g"l-h-i, control 
respectively, P > 0.05). 

heterospeclfic psirs. 

cat«9, palied birds ranged In succeas.asr Ainaziiia 
laC) > AjMsIlla (AA) > Chlorostlltion <cc) > Chiorastiibon 

slightly higher 

overnight expenditures in pairs (174.0 ±65.6 J*g'i*h"i) 
compared to controls (see Table 4-4; P > 0.D5), whereas 

141 followed the same general pattern as energy storage, wit! 
a trend of higher profitability for Amaaiiia compared to 

a again a good predictor o 


daytime energy expenditure, using 
number as factors with hover time as the 
!t for differencee emong y-intercepts (Ambs) 
elopes (incremental coats of flight;, following Tatner 
Bryant (19861 and Bitt-rtissen st al. (1989). All groups 

d Ajnagiiia {TdbXd 4-7 
-5, rdspeccively; 

R«ljtloinfiin« b 

n behavior A 

ceceived showed no sl^niflcenn cosrelaeions with any of the 

ag^teesion initiated to correlate negatively with both Intake 
and axpendicuce, with only one aignlficant c 

n expenditure iperiode 1-3, R - -0.417, p < 0.05) but i 
with daytime energy expenditure (rig. 4-15J 

period had marked energetic consequences, rectoring oi 

losers), feeder control significantly affected both ent 

category: P < 0.01), but expenditures were not reduced 

'n control o 

differences in behavioral and enerqecic variables between 
solitary and paired birds {both apecies combined ) , 
cacegorlaed by their ability to control the feeder. Compared 

increases in meal size, incerbout time, and hover time with a 
trend of increasing expenditure. This higher expenditure 
coupled with a slight decrease in intake (nonsignificant) 
resulted in a highly significant reduction in energy storage 

feeder control showed similar though moderated reaponsea, 
but without any significant differences (due in part to the 
amall number of periods in which birds tied] , in contrast. 


r«flecc differences i 
Most significant was 

eapended energy at a higher mass-apeclfic rata, a 
required e greeter mass-specific cate of lotahe, 

traplioer*s tendency to hover somewhat more . 
its higher mass-specific resting metaholic r. 
hMP for the smaller species Is eapected, basi 

adequate energy for lasting the night 
ir 24 h and without resorting to 
control birds thus sepreaent a sound 

profitability to stc 
baaellne of energy regulation 
predictions for both species were generelly eupported. 
{both species combined) . This Increase in expenditure 

to the feeder (nearly ail Chlorostlibon in ac 

ifttaae. for both control and paired bizda, ragresaiona of 
suggeat that above expenditure ratea of ca. 000 - 0SO 
than expenditure, for solitary birds, this upper llovit waa 

responded as night be expected for an exploitative conpetitor 
using pre-enptive harvesting, Interbout tine decreased 
nearly 301 conpared to controls for ChloroatiJbon in 

this increase doos not Indicate that Chioroacilbon was not 
trying to maintain a high visit rate, only that Anaziiia was 

X individuals h 

.6 posaible "ceiling'’ 

(neasuce of Che "accempca” ChioroatiJbort tnefle CO visit the 

AnaeiJls is 

: pairs (Pig. 4 -va) : about once everp 90 t 
its ccntcol incerbout Cine of one visit i 
s, Cbloroscilboa apparently greatly 

frequeccy in AC pairs by casing larger ineais, its ovecaLi 
intahe rate dropped slightly relative CO controls. Because 

energy storage in heterospecific pairs (105 J'g~l.h~l below 
controls), as predicted. In conspeoific pairs, contrary to 

storage (69 c.g*i-)i'l below controls), alC)(ough less so chan 

possible ceiling 

respectively, cnnslderably higher than the 

The results for Anaeiiia confirm most of the preOictions 
for an interference competitor. However. Amarliia in AA 
pairs reduced hover time relative to controls, contrary to 

as predicted, despite flying leas, this was due partly to 
increased FHHs and possibly to increased bat nonsignificant 
incremental costs of flight when in pairs, As expected, the 
AA birds on average had reduced energy storage, relative to 

frequency an: 

a of each pair. In conpariecn, Amarilia in 

The behavioral variables of aggression, when used to 

for winners 

r control on 

response i 

e energetically mere successful 

nazilia shoving a 

ken in ccnapecific pairs. 

Che aefensihiiity of Che feeder. Vhen a food soui 

enec^ecicaily more docinp those periods of time it 
with AiPssJlJa chan when paired with chlozostllbon. 

connterincelcive, bee 

"intruder " 

i aggression le 
a in agonisclc 

Inane birds may stay perched c 
feeder, flying only to repel 

when both wing disc loading end foraging me 
a hummingbird species, ics energetic r> 

measursble ei 

sharing a 

consequences, despite 
required Cor hovering IChspter 

parclAlly be explained by dlcCerences in mean 

predictipna b 
regulate enex 

can alter actual wine dlec loadings 

capable of initiating 
conapeclflc paita. Finally, 

icb aa net energy gain, nay net 

optiaial foraging (PeBenedletls et 

r optimal n 

uaed to fomulate ru; 
frequency, and patcb 

wben oonpetltlon la abaent or ; 
and hence affecta only food availability. However, when 

occur even when apecies are primarily exploitative 
competitors, subordinate individuals may have little control 
over the sire of each meal. Its timing, or when to enter or 

is purely eaploitatlve 

Foraging rules for subordinates cannot be derived 
empirically by measuring their actual meal sises or bout 
intervals. For example, in this study a long interbout t 
for Chlorosoilbon in AC pairs did not reflsct the rate at 

meals taken by Amaailia in AA pairs do not neoesaarily 
reflect a "deoialon" to compete exploitativaly. Instead, 

These caveats are not intended to snggest that 
humminobicds do not need to achieve a certain rate of net 
energy gain over the foraging day fcf. 'propoitional centre 

hununingbieds do not follow 

may elmply be "get vhac you can »hen you can"— a variation oj 
Che "making the bast of a bad 5ob" rule (Hainswocch and wolf 
t some individuals of both species, in both pair 
'k the largest meals they possibly could (Table 6- 

ngbirds facing competition need not adhere to the 



variable * X] ai baseline Icontrol) dayline energy budgets 
{periods 2-4) for solitary birds. Symbols and sample sizes 

Equation p 

Anazilis {AS) and Cbioroscilbon {CO in oonapeciCic (C) and 
for calculation of tbeoretical crop volume. 

Expenmemal Cage: 

Day 1; Solitary birds (rial (conirol) 

p -s 

Day Z: Birds in pair tnal (aaparimamai) 

-g ] 


s solitary Individuals separated by 
Us ot protocol. 

— •— Amaasa (AXA) 













Figure Daytime Profitability Index iPI). See cbapter 

FIg°%-3^°“^“'^^°° treatments as tor 

(per hour) 

0 20 40 SO so 100 120 

(per hour) 




B pcevlous field work OQ durmUigBlrd enereetice tdac 
;ijsed OA cercitorlal Bicda at rich nectar acurcea 

(Carpenter and MacMillan 19 
not he repreaentatlve of (no 

blrda manage t1 

.1 energy bedgete u 

availabilltiea and Interflower diatancee, thereby aimulating 
naturally occurring foraging conatralnt (defined in Chapter 
1) under conditiona of no coopetition. ay my definition. 

unlimited and transit costs (deterolned 
budgets obtained under these conditions 

Chapter 2). In contrast, birds face varying degrees of 
foraging constraint when Coed is limited or transit costs are 

topics, rirst, I Investlgsted t. 

individuaX hummingbirds often face marksd fluctuations in 

their energy budgets. 

adjustments in the terms of 

may adjust its time budget 

potential changss 

respond to conserve energy by reducing the tii 

welts longer between foreging bout8/ flight C( 
lowered and foraging profitability (defined a: 

require foraging 

rate remains constant. Intalce should then be increased to 
coispensate. if possible, if not, Chen a bird nay reduce 

possible adjustments in response to foraging constraint, with 

responses a bird exhibits. 

tested U 1 s t9T9« flight ceqe at taro levels of food 
availability (ad iibicuis and ca. 361 below ad ilbitum, HIGH 
and LOU, respectively} at eacb of two distances <4 and 20 tn 

dispersed o: 

s represented varyin? conditions of foreging ci 
Because birds were tested under conditions o 
I expected that both species would d( 

0 foraging constraint} rat}ier than widely 
w-reward flowers. Thus, each species should 
n feeders were NHAR conpared to FAH and when 
oiapared to LOH. Based on wing disc loading 

about interspecific differences. Because of its lower wing 
disc loading {and presunabiy lower flight costs) and its 
known ability to exploit widely dispersed and low-rewerd 
flowers in the field, I expected C. 
greater energetic success than astasiiia u 

s adlust energy budgets to conserve 


twees February aj 

plaecic eheetlsg (3-5 

an electronic balance ac one i 

on dn electronic b« 
stenddrd feedece vi 

kept filled. Tour identical 
e evaporative water loaa 

evaporation for tae total main feeder 
Che LOM food creacaienta, aolution was 
miorosyringe autodiepensers IRamilton 

all au&iliary 

converted to Joules ai 


cop^leced d roood-ccip flights (perth-to-fee<l6r-to-p«7c:h> , 

Individual 1 cn^ haldlng tags providsd with 20,0% sucroa« 

it was confined Co the area Burrounding the patch by a 
mosquito net curtain (Fig. 5-1) . The bird was given ad 

obtain its initial digesta-fcee mass (Chapter 3) . Ac 0741 

I recorded two behavior pactecna, flight time and nusiber 
of feeder visitSr during sis SQ-nin observation periods 

vialCB Co Che mein and auxiliary feeders were tallied 
far each observacioa period and eumned for all periods 

feeder-co-perchj - Additional vlaica to Che same feeder 
{during the same flight) were scored only when 10 3 or cnore 
elapsed between consecutiee probes (Co Dain or auxiliary 
feeder) . I recorded flight times for each observation period 

percent of flight tine necessary for gialcing feeder visits as 

ITh(Vm) •• Ta(V,) J (100%)/Tfi„ 

and auxiliary feeders, respectively, 

lmpcrc*nt functions, but It is more time chan a bicd 
absolutely needs to make its given number of visits. 

Energy budgets ware calculated gravimecrically (Chapter 
2), using Che mean digesta-fcee body mass ( [initial+finalj /3) 
for each bird during its trial. Energy Intake cate (!) was 

period (high food only! and ( 
( variable and food (LOH va. 1 

ticne exhibited a distance trend (FAR > N£AR; P - 
0.078}, with a significant interaction between food and 
distance levels (P < 0.05) indloatlng a stronger effect of 
distance at HIGH food (Fig. 5-3) . CMordSCiibon spent 
consistently more time flying than did Asrazilia across all 
treatments (57.0 ±23.5 vs. 41.9 ±19.4%, respectively; P < 
0.05) . Following visit rate, flight time increased 
dramatically at l,OH food compared to HIGH (p < 0,01) . Flight 
times varied over 40-fold among indivldgal birds, from 2.3 to 
100% IChlorOSCilbon in LOW HEAR and HIGH FAR, respectively) . 

Both species eppeaseh to engage in oonsiderebie flight 

in more o£ this extraneous flight when feeders were FAR 

between feeder distance end food level {P < 0 . 001 ) indicating 
a more pronounced effect of distance at LOM food. Birds also 

-0.744, P < 0.01), bet no significant 

suffered significantly 

to HIGH food groups, birds on LOH food 
lower (negative] storage races (p < 0 . 011 , with a trend for 
Cbiorosciibon to coecabolire energy stores faster chan 
Asaalila (P - 0.062) . Peedec dlscanae had no effect on 
energy storage. 

Within each species, birds visiting HIGH food feeders 

Chlerostiiben consuming at a slgnlficancly higher race chan 
Hmaailia (622 ±162 vs- 563 ±135 J-g~i-h"i, respectively; P 
O.DS; Pig 5-6) . For both Species combined, Che fixed level 

Jibicum) Chan the HIGH food groups (P < 0.01) . 

hates of total energy expenditure showed a significant 

more energy Chao Amaailia in all treaCoents (mean difference 

• 13.55; P • 0.072) . ToCel energy expended showed m 

directly related to flight cltne for each species, and 
species had y-lntercepcs significantly d 

♦ 4.7B (1.65) X Trt (5.2) 

< 0.001), 



where R is tocal energy expenditure end TyL is the 

percent time spent flying. Species did not differ 
slgnlflcencly in either slope or f-iotercepts (both presented 

ANCOVh model, to test significance of the flight tine x 
expenditure Interaction, indicated that all groups had 

iatercepcs, were then cong>ared using the standard AKCOVA 
model. Because the Y-inceroept is the energy expenditure at 
0% flight time, these values represent the mean resting 

s (following 

species (Fig. 5-8). 
s, having the lowest 

varied considerably, from a low of 317 ±71. l J-g-l-h"l for the 

G#ngral Rgaporsgs 

ForaQlaQ Cacutralnc 

by feeder disteece, vith no diffeceecee between species, the 
fors^ing of LOM tbod resulted in varying degrees 

energy storage be explained in ti 
flight time were dramatically higher at ton food cocnpared to 

erms of the behavioral and 

increased flight c 

on the flight titan nenand t 

decreased at FAR feeders. Plight 
distance treatment effects on visJ 

e eiaits iBqn. ' 

ilcde visiting PAR or SIGH food 
1 flying than the minimum required 
1 food groups, birds visiting the 
percentage of total flight time 
visiting NEAR feeders. Howevsr. 
tad higher total flight times than 

High flight expenditures when food was limited explain 
extreme loss of stored energy by the trapliner Chiorortilbon 

feeders, with Chlorostilbon flying more than Anazilia, yet 

nighttime torpor (Borthotooew et al. 
HaLflsworch and wolf 1970, Carpanter 1 

r RKRa to drop 9i 

usually kapt 
; aluogisb and 

Perching hirda naintained sleek plumage 
their eyes open, hut often appeared sOiiM 
low body temperaturea (judged by touch ca. 30-33 *C, 

alertness, preening, calling, and beating wings), not 
the energetic costs of sitting per se. A reduction ii 

accivlty while on the perch could reduce rhrs acd would be 

calatiuely low RMSs because ot cceatec specific dynamic 
action (3DA) in BICB food groups. Specific dynamic action is 
the increase in metabolism that accompanies digestion and 

AS groups. : 

perching bii 

which in my experiment 
food availability IHIGI 

Marloh 1919b) exhibited this response to cold ambient 

defined here, 
:h ad Jibirum 

> 80 ), CoJlbrl delpflinl and 
Schnldt-Marloh 1979a) and 
1 Sohmldt- 

limitation (Carpenter 

during Incubacion at a warm ambient temperature waa reported 
bouts at the nest these females became hypothermic (and hence 

daytime expenditures for LOW food birds were 84 lower 
(or HIGH food birds, with no effect of feeder distance, 
although both species lost mass at LOW fodd, these 
( would have bean much greater without DHC. Had LOW 

houc DHC, respectively} . 

egual to unconstrained groups {HIGH 

The energetic response of 
stretegies of several flying 
endothecmlCr maintaining high 

(Heinrich 1970, 1971, 197<, 1975, Heinrich and Bartholomew 1' 
ia energetically diaadvantageous for small animals beeaue. 

r relatively high ci 

avoid detection 

endothecmy Is abandoned (or reduced), Tss will rapidly 
to levels at which significant energy savings will be 

temperatures in Monteverde are relatively warm and see) 

reduced start-up costa (Hatunel et al. 1969). If tropical 
hunmingbi Ids do not often use DKC, it may be due primarily 
the risk of predation, perhaps only when this risk is 

foraging constraints will hummingbirds be likely to all( 
energetically under conditions of no foraging constralni 

which AmeziliB and Chlorostllbon malnCeined positive rates oC 

as Indlcacea by the 

hoitoeeneity of slopes in the preliminary AVCOVa. ThnSf onci 
sqaio interspecific Cifferences in wing disc loading did noi 

expenditure (see also Chapter S) . However, the wlnq disc 
wln^ disc loading of an Individual varies erratically over 

and such fluctuations night h 
wing disc loeding from that b 
Faotors other than wing 

•sc loading can affect 
wing disc loading. For example, energy expenditure depends 

dia«n9iCLAS and my have caused 

fly at suhapcldal velocities, compared to chose used in the 
wild, chus masfeiag diflerencas due co wlnq disc loading. 

Consiscenc with pcevioualy described foraging modee, the 
crapllner ChiorosciJbon appeared predisposed under all 
condlcions co fly more chan the territorial amarXiia, 

Beceuae greater flight time reaultsd in consistently higher 
foraging constraint. Contrary CO prediction, Che Crapllner 

low-reward crapllner requires greacer overall masa-apeclf ic 

ChlorosclJhon. In this treatmenc, Amariiia had the lowest 
vlslcation race but highest flight time for any creacnenc 
group of Chat species. As a result, Asiariiia HIGH PAH 

during LOW food. This difficulty in energy management by 


Ajn&2ilia spent inucA cii 
guacdlng tbe feedec. 

lid rddolCdcl 1 


n^ura S-1. Exparlcoerital flight tube. In N£A? feeder triela 
closed) . In FAR feeder crisis Che bird flew through all five 
see down segoencs 1 and V. See case for description of food 
delivery rates and feeder distances. 


Figuce 5-2. Visits to inain aod auxillazy feeders sonsned fi 
ail six SO-rnin observatioe periods (Fl ■ 10 per cell}, the 
tecrltorlalisc ajnaziiia {circles) and t)ie low-reward 

food and at NEAP (open aymbols) and FAS (solid synbols) 
feeders. See text Cor description of food and distance 




Figure 5-5. Dayclise energy storage rates (J.g'i'h"!) as a 
function of food level and feeder distance for AnerlJla and 
Chlorostllboa. see Figure 5-2 for description of synbola . 





storage with a nigh degree of accuracy for each particlpanr 

meals for each bird auac be recorded Imniedlateiy. In 
preliblnary expericnenta T deCarmlned chat 1 was unable Co 

were available. Thus, co conduce my final experiraenc , in 
which two birds shared five feeders, I used Che doubly 

on the Tsochilldae, only two publiahed investigations (Powers 
and Nagy 19B8; Heathers and stiles 19S9I employed the DLH 

family. The accuracy of the DLH method has been determined 
for several bird species, but prior to this study the method 

$tudie« done with other species oey not genersllze to 

conditions identical to those ondes which X would oonduct 
experisient S (Chapter 7] . 

1 conducted s specie! validation on a tropical 

id HcClintocIc (19SS1 recognized the problen oE 

equations to correct Eor these effects in studies 

2 hh 1*0 have incorporated oorrections for fractionation effects 

fracciohstion o 

1979) . The DLN studies done usinp tritium <7 h] instead of 
(Nagy 19B01 may he subject to larger fractionation effects 

Nevertheless, validstion studies on five 
Involving 3hhUo (BtO-181 with no cotreetlon for Is 
fractionation, have indicated good accuracy (mean i 
ranga from -4.91 to +6.51, range for individual va 
-171 to +161, summarised by Nagy 1989) . However, ] 
hunmlngbirds have estremely high rates of water fli 

in the body differ in isotope activity levels (and hence the 

s {Chapter 3) and glucose 

r intestinal li 

unlabeled water fcoo food may not mix completely with 
isotopically labeled body water before being excreted. 

Because of these potential sources of error in applying 
the DLK method to hurunlngblrds, coupled with increasing 
accuracy In measuring isotope concentrations that allows 
detection of small errors (Schoeller, hsltcb, and Brown 

conditions simiiar tc 
experimental design w 

5 determine the accuracy of the DLM 

sing HTO-IB with and without 
n previous experiments with the 

Haterlals n 

The use Of tritiated water was approved by the 

of Florida, Gainesville, FL, USA. Fensission to 
aearch with radioactive isotopes in Costa Rica was 
the Departameoto de Control da Radiaciones 

fed Bd ISbitunr from a prewei^hed vial (±0.001 9) of 20. Ot 

etandard fi 

*ere run concurrently Co moaaure 

initial injection z weighed Che bled again (prior Co feeding) 
and drew a blood aanple from Che jugular vein. Following the 

California, 10a Angeles, CA) for analysis of isocope 

da Aecuraos Natureles cnergia y Kinas. DirecciOn General de 


Energy budgets for both nethods were calculated Co fit 
Balance aar.hofl faap I calculated dally water Influx 

(Mvi-Mwf-Mevap) ( IBjOl • ( ISUCl 

foed (0.80 9 «iO/g lolutlsn), (SUC) is cbe 

storage vied the same qravimetclc method desetihed In Chapter 

blood sample can severely atcesa hutmalnpblrde (8. Tlebout, 

Hagy (19S3] which was recently validated for small birds 
(Webster and Weathers 1989) . I obtained initial isotope 
activity levels from six additional A. saoceroccei not used 
in the feeding trials. I injected these birds exactly as f 
the experimental birds, and drew blood samples (20-40 pi) 
after a 40-min eqvilibration period. 1 estimated total bod 

not significant. Thus, I used the mean T8W/BH (0.844) to 
estimate TB14 from the mass of each experimental bird. As a 
independent check on this TSW/8M velue, I else weighed 8 A. 

. 9<auc6riscc«i. 
i«9e equations do 

ng equation 6 

:1966) to correct for fractionation, 
tlon 32 of Lifeon and 

final tritium actlvltiea in blood aatnplea« and t ia tine 
elapsed in days), 5 ia average moles total body water E (K1 + 
V2t/Z]r and 1.Q33 is the isotope fractionacion correction 
asauning chat the physical fractionation factor for tritium 


fractionation affects 

35 for tritiataO water to yielO 

r'coj - (B/2-00) (Kiso - KjhI -O.OmiKjnHl (5,4) 

where Kiso - ln(i»Oi/i»Oj) /t, 
tlcCUntcislc (1966) at 

h and ^®Oa representing 
,n blood saniplea. i asscoed 

energy expenditure 

Rolw • rcoa fsioloa/d) (22.4 l/nole) (20.9 

CCS for fractional evaporation of 

I calculated daily energy intake froB H20st.s with 
lotw - (HjOsiw) ((SUC1 x16-49)/UH2O1*(IS0CIx0.S79) ) . 

lOLW “ Rntw. 


CAlculaced £:om cujo and rco^; corracted estimates (HsOolv:' 
ate.) wera calculatad frera r'K20 and r'coa- 

As an altarnativa to the direct DLH eatlitacas, I also 
calcalacad two indirect estimatea of expenditure and incate 
B’ntur " loluj - feat (6-8) 

”0 dilution apace and from 
differed eienificantly (T * 2.26, I 

anali over the 24-h trial period (' 

calculacions using ttie values/ because tbey were obtained 

the BAL ana DLWt (corrected 
variables these two methods 

significantly different from 
differ significantly, and foi 

both positive and negative values, calculations of pe 
S-1. For DLHz and hal methods, the F means differed 

energy Intabe os expenditure, small errors in estimating z 
and R can result in large relative errors (I) in estimating 
Potw. The unoorrectod method (Dl»i) yielded HjO Influx 

s (except P) there was also an extremely 



wexe significantly cores 
inClotlnguiahabla ataciatically 
aatlnacsB, Iol«; and Fol «2 / and ■ 

e significant lac 

1 and BU. loethoOs suggests that physical 

way be appreciable in the tropical humriingblrd species I 

fractionation using published values, the 
highly similar to gravimetric estimates for all variables. 
Thus, comparisons between studies using either the DLHg or 
gravimetric method can be made not only in relative terms 

water Influx alone suggests that water vapor input via lungs 
and shin Is cot a significant source of error (see Llfson and 
McClinCoclc 1966; Nagy and Costa 1980 for a discussion). This 
validstes tritium ss a reliable marXer for estimating water 
flux in huimDingbirda living in humid tropical environments. 
The signifiesnt correlation between both methods for water 
influx, energy intaXe, and energy expenditure suggesta no 
evidence of systematic error in either method and indlcatee a 

high degree of reliability 

eatinatea, if possible. 

rinally, Che equivalence of low, vs. laai. and a'ow, vs. 

injected with tsitiaced water {BTOl . Assuming a diet with 
precisely Kcown water and energy content, an Independent 
detertalnatlon of TBH/BM, and nininal change in body mass over 
24 n, then HsOolw,, low,, B'dlm,, and Pdlm, can be calculated 

En«c?y Buclgees 

Energy Expenditure <kJ/d) 
(Selence Method) 

Energy Storage (kJ/d) 
(Balance Method) 

(A) and energy budgets (C-P) for corrected doubly labeled 



When nectac la lipitad, botn exploicaclva and 
intacCeranca coppetitioo for food will raduce energy 
availability per individual by dividing neccar production 

flowera, regardleso of 

density, resulting in Inczaased 
consuped. Exploitative 

coppeticion, . 

transit costs above those icnpoeed by zaducad energy 

ate suffioiently dense, part or ail of a clupp nay be 

territory defenders and for intruders, due to increased 
flight time associated with chases and other aggressive 

production, ■ 

.0 confioquonces o 

or dlsporfilon, flower defensibillcy, and cbe 

f Intecapsciflc conpeticion for limited 
each of two patterne of feeder 

e to provide the most parslcoonloos 

more marked energetic effecte in response Co Interspecific 
competition for limited fooO, because ChlocOBtiilaon would 
likely experience a greater discrepancy between energy intake 

different energetic effects on the terrltosiall: 
trapliner, particularly when food was limited- 

interspecific competition, especially whan clumped feeders 

AraazlJia and CAioroatii&on occurred wnen flowers were dense. 
Under these conditions/ he found that Ameziiia had a 
relatively large conpetitive effect on the local population 

1 tested hetecospeclfic pairs in a flight ci 
feeder dispersiaa treatments. In the cluorped t 

disperaed feeders. In contrast, when feeders were dispersed 
1 expected ChlorosCiihon to fere better chan /tsieziiia, and 
better than ChiorosCilbon at clumped feeders. In addition to 

(S') and feeder overlap batween cagemstes ( 

by specieS/ by feeder dispersion, or over time. 

were arranged la a dianona 
feeders could be seen OdX^ froin perch III, fcon which they 
by a bird on perch III or In flight anywhere in aegnient 111. 


doubly labeled isotope mixture icbapter 6) . Amariila 

fllgbc tube (Pig. 7-11, where they were Kept behind the 
aih without food 
o adjust to the flight ti 

reedy to feed. At 0715 the first delivery of sucrose 
solution was made to the experimental feeders, and the 

rate was sufficient to Keep birds ac 

birds, alternating species, during 1 
periods per day, beginning at 0810, 

ile beeping ti 

followLri9 raoming, 24-h 

CAleuLate Its rn^an b> 

Agonlscic acts Initiated and received {Chapter 2) within 1 n 
of feeders were scored as feeder defenses and those farther 

Included a small proportion of miscellaneous chases and body 
contact not clearly associated with either feeders or 

received (per individual) . Because agonistic acts 

all 12 observation periods, night time and viait behavior, 

only 6 observation periods per individual for each trial, 
for each of six observation periods i also calculated 

whese Pi and ^ are Che pcapoctiona 

Energy budgeca ware calculated i 
deacribed In Chapter € and preaenCed 
by dividing time racea by each bird'e mean Body ma: 

calculated foe the perioda (a) before entering and 
leaving the flight cube, and (c) during the flight 

ITRIAI • ItOMU “ tlpRETRIAL ^ IpCSTTRlAl) f 0.3 

f viaica to each f« 

(cot^lete overlap) . 
s3.-8pecific ratea 

where Iiaiai. is energy inta)te while In the flight Cube, Iioiil 
la total lnta)ce {calculated using the 0U4 method) , and 
ienarsiAL end IposTtwat are IncaXe before and after Che flight 
tube trial, respectively {calculated gravimetrically from the 
pre-weighed food wlala, the Bal method from Chapter 6) . I 
also calculated a second estimate of energy storage using the 

tclal factor. BeDavioral ctata were also t< 

uaing InCapenCant t-testa batwaan speclaa or faaOar 
traacmanca, and paired t-teata batwaen parioda for a given 

Evan Cbaaa lower viait rataa are bigh compared to aolicary 
bisda feeding ad iibiCoai (aee Chapter 4}. Relatively high 
viait ratea. oouplad with the obaervation chat all feadara 

d during each trial. Spaclea differed li 
In magnitude and pattern over tlm> 

va. H2.6 viaitG/n for parioda 1*3, Anazilia and 
ChioroGCXf-bOdr respectively}, wltli only a relatively email 
interspecific difference during tne afternoon (63.8 vs. 44.1 

respectively} . 

CtilorostiibQn made very few visits early and late in the day, 
bet had a marhed pealr during period 4. 

defense, and total agonistic acts] ware higher for be 
species when feeders were oloinped {tieatment efiectsj 

These measures were higher In both £ei 

dispersed feeders to 

.es of total agonistic acts Initiated by 

I capable of short periods of extremely 
tn defending the clumped feeders. 

the feeder 111 times and the perch 4 times, for a total rate 
y rarely Initiated aggressive acts, 

When feeders were dispersed, total 

e day. The early morning high of ca. 18 acts/b 
lata afternoon low of under 4 aots/h. In 

ended with hi^n rates ot a^ccression {ca. 40 and 52 accs/n, 
respectively) with a significant depressicn occurring during 

Birds averaged from 21-2ie of the day in flight, without 
significant effects due to either species or feeder t 
IFlg. 7-S). Flight time decreased significantly ov< 

different rates for each feeder treatment {period x 

feeders were dispersed flight tine declined monotcnically, 

although mean overlap for ciunped feeders was slightly more 
(up 154) than for dispersed feeders tsS.S t24.54 vs, 42.2 

typical pair of )iirda sharing dlsparaed feeders, 
effects of )3oth time of day and apaciea (Flga, 7-iQ and 

respectively; i 

In general, both epecles e 
a decrease In visit diversity over time {period effect; F ■ 

decline. Both species started the day with equally high 
diversity exhibited no nain effect of feeder treatment (1.05 

periods was 

) and for 2S h, exhibited similar pactems (Fig. 
7-12) . Both species had equivalent races when feeders were 

ChiorosCiiboR hed increased iotahe (species x feeder 

having IntaJce rates 26.91 higher than Amasllla during trials 
with dispersed Cseders (617,4 1149.5 va. 486.6 tS7.0 
hd-g-l-h-l, respeetlvaly, T - 2.60, P < O.OSi tor dispersed 
feeders only) . Differences in visit rate were reflected In 

species had an effect on food consumption following trials 

svecagiog 261 more than AmesiJia (species effect: t - 22,25, 
P < 0.001). The differences between species showed a trend 

Both measures of energy storage agreed i 
and 3U. estimates indloated thst Anrasilia h. 

species maintelned slightly positive 

energy storage (O.SflS ±^.123 both apeclea 

eombiftad) . 

During 1 
diveralcy, E 

slmiXar patterns 

r overlap, and Slight 

set of feeders that were likely not visited by a 

reducing v. 
suggests t 

could occur over a short period of time simply as a result of 




behavior, which varied with apeciea. The midday peax in 
clarged coocrasca sharply with the paccern shown by the ocher 

three groups ( 

fcaguency of agonistic acta initiated by Amasllla during the 
trials with clumped feeders <Fig. T-S: solid symbols). Thus, 
Chlozoazilbon made more feeder visits during periods when 

were dispersed, there was no apparent effect of Amasilla'a 

Despite strong interspecific diffareocea In both flight- 

Aaaallia and Chlozoatilbon did not differ in flight time. 

The two speciee did. however, differ merXedly in energy 
expenditure, with consistently higher rates for the trapllner 
. This probebly reflects non-flight metaholio 

The trend for Chloroazllbon to have slightly elevated 
expenditures when feeders were dispersed does ooincide with 

expenditures . 

relationships) o 

DespiCfl ch« tPAPClous defertse of clunped feeders by 
AmaiJlia, Intrudlog CWorosClJbon were able Co malncaln rates 
of food intabe equivalent to territory defenders, 
CMoxosCiJboD apparently bad tbe greatest Intake rates during 

visits. During midday periods, Oiloroacllton also visited a 
greater diversity of feeders and overlapped tbe most with 
amarliia in feeder utUisatlon. In contrast, when visit 

When feeders were dispersed, CfUosostil^an achieved 

This success of Cbiorosciibon relative Co Amariiia, and 
relative to ChJorosciibon at clumped feeders, was probably 
due largely to reduced effectiveness by the territoriallat in 
excludiog the trapliner from feeders when they were widely 
dispersed. This increased tbe likelihood that the trapliner 

delivery before Amariiia. 

Both species had alightiy higher visit diversity when 
feeders were dispersed {up ca . 101), accompanied by a slight 
daciesse in their overlap in feeder use (down ca . 181) . 

Thus, when feeders were clumped, both species may have 
concentrated viaite on a small but similar subset of feeders 
during each period. In contrast, when feeders were 
dispersed, each species visited a greater variety of feeders 
but with less overlap. Because Cbioroetiibon malntainad 

dispec9«d, deapit? the obeetvation that ChioroatJlboe bad 
both lower visit rates sad diversity than Aneaiiia, 

food deliveries than did Amesiiia. Althoogh the trepliner 

much hidhec coat in terms of enetdy expenditure than did the 

As expected, the territorislist hmarilis obtained a 

When feeders were dispersed, the territorislist suffered 
aipniflcsntly lower intebe retea then the traplines, as 

malncelned positive 24-h eneroy storage. Because AmeaiJie 

tban flying during aggreenive 

t Chapter 4 ) . 

night during agonistic Incacactiona could be more costly 

suboptimal speeds (Pennyouick 1968, gill 1985), Including 

torpor or DKC could indicate that Amasiiia had inadequate 
daytime energy storage during the trials with dispersed 

trapliner when feeders were dispersed. 

higher Intake cates than .teaaiiia at dispersed feeders. 

h dispersed feeders, ChJoroseiibon incressed 

ates for the trapliner compared to the 



















^ Amsillis 
I I Chlorostilbon 

PERIOD 1 (.86) 

1 2 3 






n nln 


1 3 


4 5 



PERIOD 4 (0) 



PERIOD S (.«) 

n nJlfll 

3 i 
D 6 (0 

1 1 

2 3 4 5 

PERIOD 6 (0) 

2 3 4 5 


(period 1 Co 6) for one pair (AmastilB 5S and Cblorostiibon 
cube aapmeot nuoibera (Fig, i-u . Feedec overlap ItSl, Sqn. 







I s 

( i 

W lU 0.50 


g < 
§ “ 

B I 





Chapterjs 3 

s chapter 1 briefly aunmarlre the reaulea of 

experiment. Second, 1 present a few general themes that 

birda in the field. Finally. I evaluate the use of the 
mechanistic experimental approach and dlscuaa the possible 
role of energetic differencea among foraging modes in shaping 
guild compoaltion. 

Caged experimenta are by neceasity alnpliacic 
representations of natural conditions, aad their results must 
therefore be generaliaed with caution. The confidence with 
which captive studies can be generaliaed to wild birda 
depends largely upon the fidelity with whicn experimental 

which captivea behaved as they would in the wild. The former 
I address for each experiment separately in its respective 

discuss Immediately 

Two lines of evidence du^gest that captive birds 

rapidly, wild-caught humningbicds usually learned to vis 
artificial feedare vithxn a few hooss. As soon as new 

typically maintained body mas: 

ranging birds, and I observed 
captives. In the communal aviary 

Id apparently excellent 

d from and defended 

readily hawked for Insects, bathed in fresh water, called am 
displayed from perches and in flight, and "explored" their 
surroundinga . After lesa than a week in captivity, birds 
seemed relatively indifferent to my presence and exhibited 
all of the ebove behavior while observed quietly from as 

Learning probably played an important role in ahaping 
birds' bshsvior in the csge experiments. First, effects of 

prsdisposed iodividuals to certain types of responses. 

appeared inappropriace to tpe ecological conditions simulated 
In the experiments. Second, birds may nave learned novel 
behavior patterns while caged, both in the coiwnunal aviary 
and during the course of ectual trials. 1 attempted to 
control for learning effects throogh random assignment of 
treatments (when e given subject wss used in a repeated- 

for the OLH validation. Chapter d] . 

an entire foraging day. 
subsist on limited food 

depending on the experiment. Thus, food accumulated in 
feeders as a step function, as opposed to the more natural 
pattern of gradual accumulation described as a ramp function 
(sae Lucas and Maser 1969) . Conseguantly, experimental birds 
typically encountered either an empty feeder jiasc delivery 

This patcero resembles Che "bonanse- 
blanh'' nectar distribution described by Feinain9er {lilQ, 

indicates that AmasiJia and 

iemond et al. {1986) and Kacasov 
. My first experimsAt/ hovever, 

di9eative constraints and activity {fli9ht vs. perchino) is 
weaX, and digestive processes are probably not nomally 
impostanc determinants of enecgy budgets for these tec 

distance migrants. 

feeding opportunities. Because Selasphozus 

t during relatively abort feeding periods, digestive 
energy demands for wild ajeasiiia and CAiorostiibon could be 

storage tbese birds may need to process food at extremely 

birds that appeared behavloceliy to be winners we 

losers, but both paid appreciable energetic coats 
species exhibited different patterns of energetic 

relate to ttielr reapectlve foraqlng modes. 

The energetic atreas experienced 0; 

single ChlorostilLon La unlUcely Co share a flower clump with 
a territorial AjnaailiB if undefended resources are available. 
As a proximate mechanism, CMorosciibon may depart from 

raCea of energy storage, not simply because it is physically 
excluded from feeding in the patch. Other possible proxlmaca 

chough the single feeder was actively defended by Ajnaaiiia, 
ChlorescLlbof: obtained nearly as much food in hetesospecific 
pairs as it did wheo alone. la addition, despite Che high 

o partly determine where Ohiorosciibon will 
feed in the wild, with the tcapllnar tending to avoid patches 
defended by Amaaiiia because of learned o 

contributed to 

This eipetijietit, however, was heavIXy "acaoiied against" 
e trapllner by simulating eatEemely olumped cesournaa. 

only was the single feaOsr guite easy for the dominant 
smaller trapllner had no other feeding options available 

source and could fill Its crop coopletely in just a few 
seconds. The ad libicusi feeder may therefore have helped to 
counteract the limitation of feeding opportunitlea, but the 
cage environment still fsvored the terrltoriaiist . 

This bias for Amarilia ia reflected in the 
territorislist ’s consistent energetic sdvantage over the 
trapllner. Comparing daytime energy storage rates, Amazilia 
performed better than CAlorostiiboh und 

speciea euffered less energetic streaa 
Chicrostiilioh was the cagemate rather t 

defensible food source with 
natural conditions of less defensible s. 
numbers of intruders, the energetic advsncage o 

increajed Doth visit race to Isodets and raeal sise. resulting 

whila contributing to higher intahe, may also be avidanoe of 
pre-asptlve harvesting <a form of axploltatlva oostpetltlon) 
by Che terrlcorlallst . This conblnation of exploitative and 
interferance conpetition le anaioCFOus to "dafensa by 
eaploltacion" (of. ''defense by depletion," Charnov at al. 
1976, Davies and Kouston 1991, Hasar 1961, Lucas and Maser 

pcaferantislly fed fron flowers at the periphery of their 

In conspeclflc pairs, Chloioatilboo increased visit 
slightly, as expected for a pre-eH^tive harvester. 
Phaethornia superciiiosus, a hlgh-raward tcapllner, when 
competing in Che wild with conspecifics {Sill 1988, see 

ChioroatiiDon in my experiments. Perhaps Chlorostilhoa w 

United-food feedere available to visit (see Chapter 7) . 

the Gicp volume fti each apecles (510 ^L/610 (iL and 300 

foraglno effort produced no reward beyond the llmlced food 
available in tbe feedera. In the field, bowevar, even at low 
flower denalcy, additional foraging effort would normally 

and captive blrda may reapond to low food aa if it were the 

exploitative conpetitoca, they may then Increase visits and 
harvest pre-emptively — an adaptive response to competitive 
pressure (Gill 1966). Finally, experimental birds may have 
visited feedera nora frequently than the ''optimal" 1 vialt 

empty feedera aimply becausa no other food sources were 
available to chaoh. Thus, Inappropriate behavior in a 

three neotarivorea (two hummingbirds and one 

e inability o 

the Importance o 
Weathers and S' 

fU3M t.ire« and mean F® for eacB apeoles. Gi 
time varies widely axuon? ledlviduals and wicb 
(oraGlBs conditions, this lacit oJ ooctelatlon 

Weathers and Stiles emphaslte "tt 
(1969, p. 326} . Their conclusion la « 
could easily cancel expenditures due t 

t importance of energy- 

e Fig. : 

! well justified, 

t to high flight time. In 
exbihited by each 

rs. energy expenditure for 
combined. The 
lata were analysed by 

each apeoles when treatment groups w, 

ChlorosclJbcin, ceepeetlvely) , compared to analysis by species 
alone, in addition, factors such as thecmoregulatlon costs 

caution in estimating energy budgets from time budgets 
(Goldstein 1968) . 

the trapllning species, Chloeostllbon osnivecii, shered 
limited food or were far from the perch, b. 

I to foraga aa very low food density. 

1 flight tine could alao be aeed to aeeh out oioce 
psofiteble patches. 

Daytime torpoi might be a cnoce effective means fox 
reducing ijnmedlate energetic empendituces, but it presents 

birds cannot repel intruders from their territories, nor take 
advantage of the departure of a territory holder. Finally, 

activity IKenmel et al- 1968J- To offset auoh warm-up coats, 
torpor must be maintained for a fairly long period of time to 
effect a net energy savings. 

remain somewhat alert end responaive. Birds also have 
reduced warm-up costs for DHC compared to torpor; this 
further reduces risks end sllows birds to use DKC 
intermittently between feeding bouts, although DKC may not 


(UgUt. Utliough 

pattsra$ i. 

Lced for foedoe vlaita, t 
Lons that could explain ita 

territorial Intruders cr dafendeiB (P. Pelnainqer, pars, 
ti.). When feeders are NtAB, all three oan be accomplished 
Le perched or vith short flights. However, when feeders 
PAR. and hence the flight tube is longer, a bird must 

.ed and hence not needing to feed. In 
s flight was most pronounced 

flight may 


One interspecific difference appeared cj 
When foraging at dispersed Icw-rewsrd flowers, tne trapliner 
Cbiorojeilbon presunebly expsriences greater energetic 
success than the cerritcriaiiat (Feinsinger 1976) . Yet in my 
experiment, when foraging at lov-reward feeders at either 

energetic success compared to Amarijia. Does this indicate 
that Cbiorossilbon does not really hold an energetic 
advantage over Ameaiiie when vieiting dispersed low-rewerd 
flowers In the wild? wo, because the little trapliner, with 

&l 9 b«r nass-spaciflc intaJte rate chan Amaziiia 

With Ita 8CE0R9 pcetiisposiclon coward nj 
expandicura, ChlozoazilbQn does not appaar pi 

cacclcoEial Anaaiila, wlch lea lower flighc 

level, as In this experiment. Amaziiia, how« 
slgnlficanc energy challenge when attempting 
oversized cerritory or when crapllnlng very i 

1 visiting 

o maintain a 

d Piaopralon a 

eeder dispersion 

strengths a 

la or an isolated clump of 
were easily defaosUile, ao that 
during most periods Amaxilia tesccicced feeder access by 
Chlorostilbon to varying extents. Unlilce experiment 2 
(single ad libitum feeder, Cnaptec fl(, however, Chloroscilbon 

another. This occurs naturelly in the field, with intruders 
more liKely to enter a territory and obtain a meal when the 

(Feinalnger 1975, H. Tiebout, pers, obs. and P. Feinsinger, 

outside the 5‘feeder territory. Thus, the traplines had no 
choice but to intrude succeaefully or go hungry. Onder 
natural conditions, low densities of flowers would normally 
occur between defensible patches. Thus, the clumped feeders 
represented the unlihely field situetlon in which all fiovers 
are potentially defensible, under these conditions, with an 
Aisasiiie to defend each patch, chlorostiJbon should 
experience lower foraging efficiency and thus lower energy 
storage than AnasiJla, forcing the traplinet to emigrete or 
perish over a quite short period of tine. This scenario 

ChlozoatHbon i 


increaaln? Inttadar pcaaaura caused the total atandlu? 

*e cooiplax natural conditions, involving multiple species 

Increaslnsly difficult to predict. 

appsoximated natural loo density flower conditions. While 
the dispersed feeders were not detenalble, they provided looce 
enecpy chan birds might typically obtain from a single flower 

and only 4 m from the nearest perch. Thus, flight time did 
not Increase significantly (dispersed > clumped by lB%r 
P > 0.1) and averaged only 25.71, compared to 53.41 when 

tiouot of time spent in (light may halp 

discussion in Chapter S) , if birds faced 
greater interfeeder distances than were used In this 

relatively s. 


Amflzliid, vhen feeders were dlepereed oiay be due In part to 
the limited number of feeders svallabLe. Ciilorostllbon had 
somewhat higher Intake rate par feeder visit than Amarllia, 

to liraitatlon d 

t handle absolute food limitation (as opposed 
e to low food density) as well as Aamzllia. 

Chiorosciibon nay have an energetic edge over Afflarilia 
because of the amaller body mass of the trapllner, as 

Regardless of feeder dispersion, both species tended to 

day. These changes in resource utilization patterns 
presumably reduced the energetic conseguences of competition 

conditions, with many flower species available at varying 
patterns of dispersion, would these trends persist? 

ten birds will profit by 

rates, if possible, bat this nai 
decreasinp both visit diversity 

visit diversity! . However, if food is relatively scarce 
then birds may be forced to overlap dxeatly and to visit 

Furthermore, overlap and diversity 1 

number of low-reward flowers ol 
would have low visit diversity 
diversity, with low flower overlap between the 
flower dispersion, then, the foraging pattern 
,, both species reducing 

Ludes Chiorosciibon 

diversity — would not likely hold. Pstterns of overlap can 
become increasingly compler when both intra- and 
interspecific Intereetions ace considered, and the addition 


The DLH experiment demonstrated that individual birds 
can make short-term changes in visit diversity and feeder 

profitabillEy. A sinilac phenomenon has been ceporced for 
tnaplinin? hermit hummin9birdB iPfiaetho^nis supercillasuai 

visits to harvest nectar before competitors (see Chapter 5» 

visited them leas and iese frequently. Ultimately, birds 
tended to segregate among undefended feeding eitea, vith 
exclusive use apparently resulting from previously successful 
pre-emptive harvesting. This process of dividing resources, 
and ultimately reducing costs, represents another means of 

trapllnecs . 

The effects on foraging birds of tbe 

dispersion; were often dependent upon context, for example, 
the advantage of a givan foraging mode depended upon food 

addition, there were many significant statistical 

significant three-way Interaction among speciea, conpetitlon 
type, and number of birds sharing the feeder (Chapter 4) . 
Such complexity maxes it difficult, and of questionable 

a sharing a food source. When food is shared, regardless of 
pe, ic nmst he divided among pssticipantsr yet 
me, sharing food tends to inorease the 
nda of pertiolpants by increasing expenditures, 
demand for food Interacte with food 
availability In situation-specific ways to determine whether 
8 bird can meet its daytime energy storage requirements. Pot 

energetic d 

example, when the shared food source le defensible, a 
resident AnaaJiia that limits access by intruders may be able 
to increase intake rates aufflciently to meet increased 
expenditures, thus ensuring adequate daytime storage. This 
energetic saccess can occur when defending against hetero- or 
Gonspeciflcs, and at abundant or limited food. In contrast, 
both the trapuner CftlorosCllbon end aubordlnate Amesliia can 
acbiava energetic success only when paired with Chloroatilhaa 

invariant and persist whether birds were tested in pairs or 
alons. The most conspicuous is thst Chicrosciiboo typically 
exhibits both higher BMBs and higher rates of overall energy 

s previously dismissed, 

Che trapliner, as pcedicced 

activity (primarily flight) . The effects of BMS aed activicy 
are lioked via positive feedheclc, because a high BMR requires 
high food intelte. By specialising on low-reward and 
dispersed flowers, OilorostiJbo/: may typically maincsin low 

relative to the territorialist. Because it has a very low 
energy storage capacity (Paladino 1989), ChJorostiibon, like 

o spend in foraging flight, selection shoold f. 

apparent in the relatively low wing disc loading of 

(all hovering) and a large cage (mixed forward flight and 
hovering) failed to detect a significant energetic difference 
between species Chat is attributable to differences in flight 

presented in Chapter 5. However, even if Chiorostiihon did 
exhibit somewhat lower flight costs per unit time than 
Amariiia, the interspecific differences in these costs still 

Additional studios sr« 
and cloasr monitoring o. 

could bs rsndsrsd relatively unimportant ii 

energetic coneequencee of wing disc loading, which nay be 
important over ecological and evolutionary time (Felnslnger 
and Chaplin 1975, Pelosinger and Colwell 197B, Felnslnger et 

Finally, birds tend to dinlDlsh activity and 
expenditures over the course of a foraging day, despite the 
constant availability of food {whether ad libitom or 
limited) . When food is shared, this reduction may reflect 
partitioning of limited food resources, which allows greater 
profitability {Chapter 7), but reduction in activity and 
expenditures also occurs when food la neither limited nor 
shared (Chapter 5) . Mild Amasllia and ChloiesclIBoe may 
exhibit a sinvilar temporal response as an adaptation to 
natural patterns of nectar availability at the Honteverde 
site (P. feinslnger, pars- comm. I . Many of the commonly 
visited flower species secrete the moot nectar in the 
morning, tapering off in the afternoon {Feinslnger 1976, 

197BJ . Reduced activity in the afternoon has also been noted 
for hummingbirds st other neotroplcsi sites {Stiles 197S) and 
may in fact bo typical for most bird species {Kendeigh et el. 

food avallebillcy, reduced competition due to pertltioning of 

TK» mehunlitlg >Bora»gh to und«r«r»ndlng 


hujnioin^blrds. Eaoh factor vas examined in one o 

axcrapolaca with any confldanca from chase types of 
axparljsanca to natural conditiona would first require a 
series of additional expecimants . Ajoariiia and ChlorostilDon 
need to be tested under conditions involving a Eoora conplete 
ranqa of feeding opportunities deaipned to simulate the full 

very large flight cages oould provide birds with both clumped 

the influence of seasonal clisiatic changes that may influence 
January each year may be more stressful energetically to 
The results of my experiments indicate that the most 

intake:expenclicure ratic 

Che domeatic: pigaoa, coZumba 2 ivie. rood cosca ware 
indiracaly contooLlad by cequiriog a apecified number of 
pecks before preaencacion of a food bopper (Paihocte and 
Henderson 1988, Osthein and Rautenberg 1959, Raahocse et al. 
1989) . Mthouqh cbese particular erperiaienta ware performed 

{Kursh 1980, 1984; Collier and Bovee-Colller 1981, Collier 
more natural axpariments (Houston and McHamara 1989) . For 

effort] , When food la limited by the amount of energy 
required to obtain it, rather than absolutely (as In my 

birds with low foraqinq efficiency (sucQ as 
) should aobleva qreater dally storage by 
foraging longer. 9uch oxperlmente may ultimately reveal 

', not elmply on their ability to tolerate 

e difficulties in generalising to the natural 

1 habitats where ^jcazllia ana ChlorosCilbon ar 
y may regularly Intecaot with 10 eSdltlonal 
d species and irregularly with yet another six 
(feinsinger 1976, H. Tiebout unpublished data) . The oche 
species utilize a variety of foraging modes in addition t' 
territoriality and low-reward traplining. These other 
species also span a wider range of body size and bill 
morphology than do Amasilia and Chlorosciihon, with all o 

d Plowright 1985) . 

resources, and Chlocostilbon exhibited 

behavior can vary altitudinally within a single hummingbird 
species, possibly in response to energetic factors {Berger 
1974a, 1974b; relnsinger et al- 1979) . similarly, at a 
single site the foraging mode of an individual bird may be 
somewhat context-epeclfic, depending upon food availability 
and dispersion, and qualitative and quantitative 

experiments Amaz. 
dispersed, limit) 

and competition 

e not rigidly cas 

characteristics of competitor prsesure (Tei 
Feioeinget aod Colwell 197S) . 

capabilities o 

e lepcrtance of 

t species-specific 

1960) have been little studied. 

constituents, such as proteins and lipids (Ba)cer and Ba)cer 
1975), reioains virtually unstudied. 

Finally, as toentioned In Chapter 1, energetic success is 

events occurring elsewhere. Birds are highly vagile and can 

c feeding 

Despite the many difficulties in eKtrepolsting free 
sifflplletic expeclitents to netursl cofoeunities, such effort: 
nay ultinately yield greater prediotive poeer {Tllnan 19871 
than the currently popular phenonenologlcal approach to 
species Inceractiooa . Furthernore, such experiments can 

above (Schoener 1998, Tllnan 1987| , The four major 

assessing t 
energetic a 

by digestive ocnstceints. Thus, limits to IntaXe In the wild 
must normally be imposed ecologically, through either limited 

ecological limits suggests that differences in the ability of 

variation in foraging modes, can have energetic consequences. 

Differencial energetic success was demonstrated in the 
last three experiments I conducted with Amariiis and 

relative foraging profitabilities and in the flexibility of 
their energy budgets, in response to the ecological factors £ 
manipulated. Thus, species representing two foraging modes 


f guild dynaioica. 

interapecific anezgetic d 
among foraging modes as groups. Because my study tested only 

trapliner, I cannot 
Second, a link between energetic differences among species 
and their consequent representation in a local guild must be 
documented uoder natural conditions with free-ranging birds, 
3y inveatigating this energetic mechanism in a diversity nf 
other guildmates, as wall aa other mechanisms importsnt to 

processes responsible for the highly dynamic struoture of 

;.;. ss.s. isis^^i-:- 




“■Eis3i‘.iSi3;3".33;."S;“.ssr‘-' " “■ 



rsss- s^ss- 

"3:;.™i43‘ 3iK..r’,3;;3;r3;?'r.",i. 


"•S3^.; s.;33333‘i;:. ssss;/53s^ 

Garland Pi4sa, Haw ^ort. ^ ^ 


■'nr-^k°..i ,;K..s:.K‘fiu^ 


.rta.;i.r;« ist4K-“" ■-” 

— i/K’‘E;:.Kin.‘." ^ 



”lSs:-:; “■ 


“'is; iii sirS'SSisi' jfi 

. Lri„S".r'. 

;‘;s; it^zixti.. 

“■:“.;i.;;.;‘S'.s;.*sf.;4K “ ss,;- •' 
"|4U„;« Tc4.4„rsar4' 4= 

■;r„;‘“ i:«is»“ - 



^H.|SHH£ESria“ = 





«icsi4n. American Midland Hati;tallse 104: 328-333. 

^mmirctlr?^«Lrernf’frrc=!Li'«rf"nLa^^ '' 

Biotrolica 10. ?94-210. 

srSSSH-ESI=sHS?" ” 



;s.. „. 



*;,ss; ;lsss."!s;s*,s a.. 

— ;SS.!.:-CSS5.,‘Ei4aa!.SK5' " 


al ot Animal Ecology 15; 319-37S. 


Harry Morgan Tie! 
Tiabout, Jr-, OQ Novel 

watchin9 animals in ci 

unOergradcace be tooJc 
changed hia destiny, 

master's bypass project ins 

Lva In the Audubon Society and the 
Jell as Boy Scouts and the Explorers. 
:he long hours he has spent indoors 

t courses in ethology which forever 

ecology and ethology. In 19B2 Harry 
University of Florida. Hia 

s thwarted by a totally premeditated switch 

Somewhat wiser but infested, he returned Co Gainesville 2 

nearly all excuses to engage in his favorite pursuit, whi. 

hihlng, canoeing, camping, nature photography, and bird- 
watching are the most oft-atated alibis. 


'rxiL^r... J...